There are five words every Vikings fan needs to recite this week, five words that explain past failures, five words that every NFL scout knows and too many general managers forget every spring:

Need is a terrible evaluator.

For many NFL fans, that phrase has become a cliche, but it is the kind of phrase that becomes a cliche because it is so true.

For those who haven't heard it, it means that drafting a player because he can fill your most obvious needs often leads to mistakes.

Vikings history is proof.

Take 1998.

The '98 Vikings did not need a receiver. The '97 Vikings featured three excellent skill-position players who produced 1,000 or more yards — running back Robert Smith and receivers Cris Carter and Jake Reed.

The '97 Vikings were a good offensive team and a poor defensive team. If they had drafted for need, they might have taken the highest-rated cornerback available with the 21st pick, and that might have been cornerback R.W. McQuarters, who had a solid NFL career.

But R.W. McQuarters was not Randy Moss.

Because of character concerns, Moss, perhaps the most talented receiver ever to play the game, fell to the Vikings at the 21st pick. And they took him, wisely deciding that you don't draft a pretty good player to fill a role when you can get a great player who will transform your team. This is the lesson that NFL teams often forget they have learned.

Take 2005.

The Vikings had traded Moss for a linebacker and the seventh pick in the draft. This time, they decided to use their pick for an obvious need — replacing Moss.

So they used the seventh pick in the draft on the fastest productive college receiver available, Troy Williamson.

One problem: Williamson was not good. So the Vikings passed on DeMarcus Ware, among others, for a lousy receiver, because they were determined to draft a receiver.

Take 2007.

The Vikings did not need a running back. Chester Taylor had easily been their best skill-position player in 2006, rushing for 1,216 yards in 15 games.

The Vikings were picking seventh again, and Adrian Peterson fell into their laps, and instead of taking a wide receiver to make up for the trade of Moss and the failure of Williamson, they took Peterson.

If they had taken a receiver, it likely would have been Ted Ginn, a pretty-good-never-great player. Instead, they landed one of the best backs in NFL history.

Take 2011.

Brett Favre was gone, so was Brad Childress, and the Vikings were determined to build an old-school offense around Peterson.

They needed a quarterback. So they donned rose-colored glasses and talked themselves into believing that they could take a franchise quarterback with the 12th pick in the draft.

They chose Christian Ponder. They could have had Cam Jordan, the star defensive end and son of Vikings standout Steve Jordan, but they wanted a quarterback in the worst way, and they got a quarterback in the worst way — by drafting out of desperation instead of clear-eyed evaluation.

The problem with drafting for need is that need tends to be blinding.

Take a mediocre player at a position of need, and you waste the draft pick, invest too much time and money in a lesser player and still, eventually, have to find a replacement.

Vikings General Manager Rick Spielman has done well when he has hunted superior athletes, like Peterson, Percy Harvin, Harrison Smith and Xavier Rhodes. He has stumbled when drafting to fill an obvious need.

This draft sets up well for Spielman. He has three obvious needs — left tackle, cornerback and receiver. He should have multiple options with the 22nd and 25th picks to fill two of those needs.

When Spielman was more forthcoming with media not aligned with the NFL, he used to hold pre-draft clinics, explaining his process.

He and his staff would group players into tiers. So if they needed a left tackle and a left tackle ranked in the same tier as their other top-rated players available, they'd take the left tackle, but if there was a superior talent at another position available, they'd take the best talent available.

Let's hope he has studied his own history, and come to understand that, yes, need is a terrible evaluator.

Jim Souhan's podcast can be heard at TalkNorth.com. On Twitter: @SouhanStrib. • jsouhan@startribune.com